Examples with StorageStripeMetadataSource com.facebook.presto.orc.StorageStripeMetadataSource used on opensource projects

Example 6 with StorageStripeMetadataSource

use of com.facebook.presto.orc.StorageStripeMetadataSource in project presto by prestodb.

the class TestHiveFileFormats method testDwrfOptimizedWriter.

@Test(dataProvider = "rowCount")
public void testDwrfOptimizedWriter(int rowCount) throws Exception {
    TestingConnectorSession session = new TestingConnectorSession(new HiveSessionProperties(new HiveClientConfig().setOrcOptimizedWriterEnabled(true).setOrcWriterValidationPercentage(100.0), new OrcFileWriterConfig(), new ParquetFileWriterConfig(), new CacheConfig()).getSessionProperties());
    // DWRF does not support modern Hive types
    // A Presto page can not contain a map with null keys, so a page based writer can not write null keys
    List<TestColumn> testColumns = TEST_COLUMNS.stream().filter(testColumn -> !hasType(testColumn.getObjectInspector(), PrimitiveCategory.DATE, PrimitiveCategory.VARCHAR, PrimitiveCategory.CHAR, PrimitiveCategory.DECIMAL)).filter(testColumn -> !testColumn.getName().equals("t_map_null_key") && !testColumn.getName().equals("t_map_null_key_complex_value") && !testColumn.getName().equals("t_map_null_key_complex_key_value")).collect(toList());
    assertThatFileFormat(DWRF).withColumns(testColumns).withRowsCount(rowCount).withSession(session).withFileWriterFactory(new OrcFileWriterFactory(HDFS_ENVIRONMENT, new OutputStreamDataSinkFactory(), FUNCTION_AND_TYPE_MANAGER, new NodeVersion("test"), HIVE_STORAGE_TIME_ZONE, STATS, new OrcFileWriterConfig(), NO_ENCRYPTION)).isReadableByRecordCursor(new GenericHiveRecordCursorProvider(HDFS_ENVIRONMENT)).isReadableByPageSource(new DwrfBatchPageSourceFactory(FUNCTION_AND_TYPE_MANAGER, FUNCTION_RESOLUTION, HIVE_CLIENT_CONFIG, HDFS_ENVIRONMENT, STATS, new StorageOrcFileTailSource(), StripeMetadataSourceFactory.of(new StorageStripeMetadataSource()), NO_ENCRYPTION));
}

Example 7 with StorageStripeMetadataSource

use of com.facebook.presto.orc.StorageStripeMetadataSource in project presto by prestodb.

the class TestHiveFileFormats method testOrcOptimizedWriter.

@Test(dataProvider = "rowCount")
public void testOrcOptimizedWriter(int rowCount) throws Exception {
    TestingConnectorSession session = new TestingConnectorSession(new HiveSessionProperties(new HiveClientConfig().setOrcOptimizedWriterEnabled(true).setOrcWriterValidationPercentage(100.0), new OrcFileWriterConfig(), new ParquetFileWriterConfig(), new CacheConfig()).getSessionProperties());
    // A Presto page can not contain a map with null keys, so a page based writer can not write null keys
    List<TestColumn> testColumns = TEST_COLUMNS.stream().filter(testColumn -> !testColumn.getName().equals("t_map_null_key") && !testColumn.getName().equals("t_map_null_key_complex_value") && !testColumn.getName().equals("t_map_null_key_complex_key_value")).collect(toList());
    assertThatFileFormat(ORC).withColumns(testColumns).withRowsCount(rowCount).withSession(session).withFileWriterFactory(new OrcFileWriterFactory(HDFS_ENVIRONMENT, new OutputStreamDataSinkFactory(), FUNCTION_AND_TYPE_MANAGER, new NodeVersion("test"), HIVE_STORAGE_TIME_ZONE, STATS, new OrcFileWriterConfig(), NO_ENCRYPTION)).isReadableByRecordCursor(new GenericHiveRecordCursorProvider(HDFS_ENVIRONMENT)).isReadableByPageSource(new OrcBatchPageSourceFactory(FUNCTION_AND_TYPE_MANAGER, FUNCTION_RESOLUTION, false, HDFS_ENVIRONMENT, STATS, 100, new StorageOrcFileTailSource(), StripeMetadataSourceFactory.of(new StorageStripeMetadataSource())));
}

Example 8 with StorageStripeMetadataSource

use of com.facebook.presto.orc.StorageStripeMetadataSource in project presto by prestodb.

the class TestHiveFileFormats method testTruncateVarcharColumn.

@Test
public void testTruncateVarcharColumn() throws Exception {
    TestColumn writeColumn = new TestColumn("varchar_column", getPrimitiveJavaObjectInspector(new VarcharTypeInfo(4)), new HiveVarchar("test", 4), utf8Slice("test"));
    TestColumn readColumn = new TestColumn("varchar_column", getPrimitiveJavaObjectInspector(new VarcharTypeInfo(3)), new HiveVarchar("tes", 3), utf8Slice("tes"));
    assertThatFileFormat(RCTEXT).withWriteColumns(ImmutableList.of(writeColumn)).withReadColumns(ImmutableList.of(readColumn)).isReadableByPageSource(new RcFilePageSourceFactory(FUNCTION_AND_TYPE_MANAGER, HDFS_ENVIRONMENT, STATS)).isReadableByRecordCursor(new GenericHiveRecordCursorProvider(HDFS_ENVIRONMENT));
    assertThatFileFormat(RCBINARY).withWriteColumns(ImmutableList.of(writeColumn)).withReadColumns(ImmutableList.of(readColumn)).isReadableByPageSource(new RcFilePageSourceFactory(FUNCTION_AND_TYPE_MANAGER, HDFS_ENVIRONMENT, STATS)).isReadableByRecordCursor(new GenericHiveRecordCursorProvider(HDFS_ENVIRONMENT));
    assertThatFileFormat(ORC).withWriteColumns(ImmutableList.of(writeColumn)).withReadColumns(ImmutableList.of(readColumn)).isReadableByPageSource(new OrcBatchPageSourceFactory(FUNCTION_AND_TYPE_MANAGER, FUNCTION_RESOLUTION, false, HDFS_ENVIRONMENT, STATS, 100, new StorageOrcFileTailSource(), StripeMetadataSourceFactory.of(new StorageStripeMetadataSource())));
    assertThatFileFormat(PARQUET).withWriteColumns(ImmutableList.of(writeColumn)).withReadColumns(ImmutableList.of(readColumn)).withSession(parquetPageSourceSession).isReadableByPageSource(new ParquetPageSourceFactory(FUNCTION_AND_TYPE_MANAGER, FUNCTION_RESOLUTION, HDFS_ENVIRONMENT, STATS, METADATA_READER));
    assertThatFileFormat(AVRO).withWriteColumns(ImmutableList.of(writeColumn)).withReadColumns(ImmutableList.of(readColumn)).isReadableByRecordCursor(new GenericHiveRecordCursorProvider(HDFS_ENVIRONMENT));
    assertThatFileFormat(SEQUENCEFILE).withWriteColumns(ImmutableList.of(writeColumn)).withReadColumns(ImmutableList.of(readColumn)).isReadableByRecordCursor(new GenericHiveRecordCursorProvider(HDFS_ENVIRONMENT));
    assertThatFileFormat(TEXTFILE).withWriteColumns(ImmutableList.of(writeColumn)).withReadColumns(ImmutableList.of(readColumn)).isReadableByRecordCursor(new GenericHiveRecordCursorProvider(HDFS_ENVIRONMENT));
}

Example 9 with StorageStripeMetadataSource

use of com.facebook.presto.orc.StorageStripeMetadataSource in project presto by prestodb.

the class TestHiveFileFormats method testOrcUseColumnNames.

@Test(dataProvider = "rowCount")
public void testOrcUseColumnNames(int rowCount) throws Exception {
    TestingConnectorSession session = new TestingConnectorSession(new HiveSessionProperties(new HiveClientConfig(), new OrcFileWriterConfig(), new ParquetFileWriterConfig(), new CacheConfig()).getSessionProperties());
    assertThatFileFormat(ORC).withWriteColumns(TEST_COLUMNS).withRowsCount(rowCount).withReadColumns(Lists.reverse(TEST_COLUMNS)).withSession(session).isReadableByPageSource(new OrcBatchPageSourceFactory(FUNCTION_AND_TYPE_MANAGER, FUNCTION_RESOLUTION, true, HDFS_ENVIRONMENT, STATS, 100, new StorageOrcFileTailSource(), StripeMetadataSourceFactory.of(new StorageStripeMetadataSource())));
}

Example 10 with StorageStripeMetadataSource

use of com.facebook.presto.orc.StorageStripeMetadataSource in project presto by prestodb.

the class HiveClientModule method createStripeMetadataSourceFactory.

@Singleton
@Provides
public StripeMetadataSourceFactory createStripeMetadataSourceFactory(OrcCacheConfig orcCacheConfig, MBeanExporter exporter) {
    StripeMetadataSource stripeMetadataSource = new StorageStripeMetadataSource();
    if (orcCacheConfig.isStripeMetadataCacheEnabled()) {
        Cache<StripeId, Slice> footerCache = CacheBuilder.newBuilder().maximumWeight(orcCacheConfig.getStripeFooterCacheSize().toBytes()).weigher((id, footer) -> toIntExact(((Slice) footer).getRetainedSize())).expireAfterAccess(orcCacheConfig.getStripeFooterCacheTtlSinceLastAccess().toMillis(), MILLISECONDS).recordStats().build();
        Cache<StripeStreamId, Slice> streamCache = CacheBuilder.newBuilder().maximumWeight(orcCacheConfig.getStripeStreamCacheSize().toBytes()).weigher((id, stream) -> toIntExact(((Slice) stream).getRetainedSize())).expireAfterAccess(orcCacheConfig.getStripeStreamCacheTtlSinceLastAccess().toMillis(), MILLISECONDS).recordStats().build();
        CacheStatsMBean footerCacheStatsMBean = new CacheStatsMBean(footerCache);
        CacheStatsMBean streamCacheStatsMBean = new CacheStatsMBean(streamCache);
        exporter.export(generatedNameOf(CacheStatsMBean.class, connectorId + "_StripeFooter"), footerCacheStatsMBean);
        exporter.export(generatedNameOf(CacheStatsMBean.class, connectorId + "_StripeStream"), streamCacheStatsMBean);
        Optional<Cache<StripeStreamId, List<RowGroupIndex>>> rowGroupIndexCache = Optional.empty();
        if (orcCacheConfig.isRowGroupIndexCacheEnabled()) {
            rowGroupIndexCache = Optional.of(CacheBuilder.newBuilder().maximumWeight(orcCacheConfig.getRowGroupIndexCacheSize().toBytes()).weigher((id, rowGroupIndices) -> toIntExact(((List<RowGroupIndex>) rowGroupIndices).stream().mapToLong(RowGroupIndex::getRetainedSizeInBytes).sum())).expireAfterAccess(orcCacheConfig.getStripeStreamCacheTtlSinceLastAccess().toMillis(), MILLISECONDS).recordStats().build());
            CacheStatsMBean rowGroupIndexCacheStatsMBean = new CacheStatsMBean(rowGroupIndexCache.get());
            exporter.export(generatedNameOf(CacheStatsMBean.class, connectorId + "_StripeStreamRowGroupIndex"), rowGroupIndexCacheStatsMBean);
        }
        stripeMetadataSource = new CachingStripeMetadataSource(stripeMetadataSource, footerCache, streamCache, rowGroupIndexCache);
    }
    StripeMetadataSourceFactory factory = StripeMetadataSourceFactory.of(stripeMetadataSource);
    if (orcCacheConfig.isDwrfStripeCacheEnabled()) {
        factory = new DwrfAwareStripeMetadataSourceFactory(factory);
    }
    return factory;
}